HUMIDITY AND TEMPERATURE CORRECTION FACTORS FOR NOX EMISSIONS FROM SPARK IGNITED ENGINES
FINAL REPORT
SwRI® Project No. 03.10038
Prepared for
ENVIRON International Corporation
101 Rowland Way, Suite 220
Novato, CA 94945-5010
October 2003
SOUTHWEST RESEARCH INSTITUTE®
SAN ANTONIO HOUSTON
DETROITWASHINGTON, DC
HUMIDITY AND TEMPERATURE CORRECTION FACTORS FOR NOX EMISSIONS FROM SPARK IGNITED ENGINES
FINAL REPORT
SwRI Project No. 03.10038
This report must be reproduced in full, unless SwRI approves a summary
or abridgement.
Prepared for:
ENVIRON International Corporation
101 Rowland Way, Suite 220
Novato, CA 94945-5010
Prepared by:
Jess W. Gingrich, Timothy J. Callahan, and Lee G. Dodge
Southwest Research Institute
6220 Culebra Road
San Antonio, Texas
October 2003
Approved:
______
Bruce B. Bykowski, Director
Department of Engine and Emissions Research
Engine, Emissions, and Vehicle Research Division
TABLE OF CONTENTS
EXECUTIVE SUMMARY......
1.0BACKGROUND......
2.0OBJECTIVE......
3.0APPROACH......
3.1Existing NOx Correction Procedures......
3.1.1Standardized Corrections......
3.1.2EPA Emission Inventory Models......
3.1.3CARB Emission Inventory Model......
3.1.4Specialized Corrections......
3.1.5Comments on Current Correction Practices......
3.1.6.Effect of Air Conditioning Loads on Humidity and Temperature Correction Factors......
4.0Recommended Practices......
4.1Heavy-Duty On-Road and Off-Road Vehicles/Engines......
4.2Light-Duty Vehicles......
4.3Small Off-Road Engines......
5.0Summary......
6.0ACKNOWLEDGEMENTS......
7.0REFERENCES......
appendix a......
ALAMO_ENGINE COMPUTER MODEL......
CYCLE SIMULATION SUBMODEL......
UNBURNED AND BURNED GAS TEMPERATURES......
NITRIC OXIDES EMISSIONS MODEL......
appendix B......
DERIVATION OF CORRECTION EQUATIONS......
EXECUTIVE SUMMARY
All of the current humidity correction factors for NOx were found to be based on historical data taken in 1971 and 1972. Some of the engines today are more technically advanced than those engines, incorporating port or throttle-body fuel injection, air-fuel ratio feedback, exhaust aftertreatment, and knock detection. While many off-road vehicles do not have all of these features, this technology is becoming more prevalent in those engines as well. The analysis conducted for this project indicated that the historical correction factors do not adequately account for operating cycles with higher load factors, or advanced technologies such as A/F control and knock detection. No engine test data were found documenting humidity effects for these additional variables. Therefore, the recommendations given here were based on the correlations developed from engine tests conducted in the early 1970’s, and the slopes for those correlations were adjusted based on engine modeling results that addressed the effect of higher load factors, A/F controlled to a constant value, and A/F fixed at a different value from the earlier tests.
The model results showed these effects to be significant and the results were used to modify the historical correction procedures. If a more rigorous approach is desired, SwRI would recommend engine testing to quantify the effects for different engine/vehicle classes.
The recommended equation to adjust standardized emissions for carbureted heavy-duty on-road or off-road (above 19kW) engines under non-standard inlet air conditions takes the following form:
Where:
T = Temperature of the inlet air [oC]
H =Absolute humidity of the inlet air [g of H2O/kg of dry air]
For heavy-duty on-road or off-road (above 19kW) spark-ignition enginesthat use a 3-way catalyst (A/F control, typically with port fuel injectors), the recommended NOx correction equation is as follows:
with no correction for ambient temperature.
For light-duty, spark-ignition engines, the recommended practice is whatever procedure is used in Mobile 6, which can be approximated by Equation 4.
Where:
Ha = Absolute humidity of the inlet air [grains/lb]
For small off-road, spark-ignition engines (< 19kW), the recommended practice is,
Where:
AFR = Air-fuel ratio of the engine
ω= Absolute humidity of the inlet air [kg/kg]
1.0BACKGROUND
Emission regulations continue to place additional restrictions on urban areas trying to achieve ambient air quality standards. Although ambient air quality standards are national, achieving the standards is a regional problem delegated to the states. However, the certification procedures for on-road and off-road spark-ignited engines are standardized without regard for regional variation in ambient conditions like temperature and humidity. As early as 1970(1), it was recognized that the concentration of oxides of nitrogen (NOx) in engine exhaust is significantly affected by the thermodynamic conditions of the intake air. Specifically, the intake air temperature and humidity have the dominant effects(1)(2)(3). Because of these sensitivities, it is reasonable to assume regional variations in temperature and humidity can significantly impact engine-out emission levels. Emissions inventory models such as the Environmental Protection Agency’s (EPA) MOBILE and NONROAD(4)(5)(6) have been developed to account for pollutants attributed to both on-road and off-road mobile sources. These models use local information to adjust the inventory based on average regional temperature and humidity for specific categories of engines.
Historically, the impact of ambient temperature and humidity on emissions was of interest because it was difficult to make comparisons of the NOx emissions from engines tested at different locations and/or with variations in the ambient conditions. In an effort to allow these day-to-day and location-to-location comparisons, various correction factors have been developed. The goal for all of these correction factors is to standardize the NOx emissions back to selected standard reference conditions, or to provide an adjustment to the emissions inventory models enabling a more accurate prediction of ambient air quality.
In light of the pressure on states and urban areas for implementing and achieving air quality standards, it seems appropriate to account for regional differences imposed by prevailing ambient conditions. Of particular interest is the impact of ambient conditions on oxides of nitrogen NOx, a major contributor to ambient air ozone levels.
2.0OBJECTIVE
The objectives of this project were to review existing data and correction procedures for adjusting spark-ignited Otto-cycle engine NOx levels for ambient temperature and humidity, and to assess the applicability of these procedures for a number of different mobile sources.
1
3.0APPROACH
The existing procedures for correcting NOx emission levels during standardized tests for ambient temperature and humidity were reviewed, along with the original reference work that developed these procedures. The correction procedures were compared to each other and to accepted engine performance and emission models for quantitative effects. Recommendations were then made on the application of the correction factors to the engine subcategories. For the purpose of this text, the main category should be considered spark-ignited, Otto-cycle engines containing the subcategories: light-duty vehicles and engines, heavy-duty vehicles and engines, and off-road engines.
3.1Existing NOx Correction Procedures
A survey of standardized procedures found two methods for correcting ambient humidity and temperature during engine and vehicle tests. The first one is for heavy-duty engines, and the second is another used for light-duty on-road sources as well as off-road mobile sources such as recreational, small off-road, and marine SI engines. Current emissions models such as MOBILE6 and the model developed through the California Air Resource Board (CARB), EMFAC2002, were also explored to identify alternative methods currently in use to correct NOx for ambient temperature and humidity. Other correction algorithms have been developed for specialized cases, though not found in a standardized procedure.
3.1.1Standardized Corrections
The EPA has promulgated the following correction factor (KH in English units and KHSI in SI units) for NOx based on ambient humidity in multiple sections of CFR Title 40(7). The correction factor is based on work performed by Manos in 1973(2):
Where:
H = Absolute humidity of the inlet air [grains H2O/pound dry air]
HSI = Absolute humidity of the inlet air [grams H2O/kg dry air]
The standard absolute humidity for the EPA is 75 grains/lb or 10.71 g/kg. These equations, in some form, are used in:
- CFR Title 40 §86.144-94 for 1977 and later model year light-duty vehicles
- CFR Title 40 §86.1342-90 for transient tests on Otto-cycle light-duty engines
- CFR Title 40 §90.419 for small spark ignited off-road engines below 19kW
- CFR Title 40 §91.419 for marine spark-ignited engines
- CFR Title 40 §1051.501 for off-highway vehicles including ATV’s and snowmobiles.
While the equation is consistent throughout Title 40, it should be noted that the use of the correction factor is not defined uniformly. In some instances KH is defined as a multiplicative correction factor to the NO concentration, while other sections define KH as the correction for the humidity effects on NO2 formation. However, in practice, the applications of these correction factors have all been applied to the total NOx emission numbers. Equation 1 has also been incorporated into SAE J1088, a test procedure for measuring gaseous emissions from small utility engines(8), and in the Texas Natural Resource Conservation Commission Technical Analysis Division specifications for vehicle exhaust gas analyzer systems(9). CARB has also uses this equation in their exhaust emissions standards and test procedures for 2001 model year and later spark-ignited marine engines(10) and small off-road engines(11). All of the previous procedures define KH be set to 1 for two-stroke-engines. This definition is not explained in the CFR, but Brereton and Bertrand explain that carbureted two-stroke handheld engines are particularly hard to characterize from a regulator perspective because of the erratic dependence exhaust emissions have on ambient temperature and humidity(12).
The original work performed by Manos tested eight vehicles based on the Federal Register Volume 30, Number 108. These vehicles were selected to represent the various engine configurations and carburetion systems found in the United States at that time. During the tests, humidity was varied from 20 to 180 grains of H2O per pound of dry air (~2.85 to 25.2 g/kg); however, the regression analysis excluded the data above 120 grains per pound (~17.2 g/kg). The temperature range was determined in accordance with the Federal Register to be between 68 and 86 F.
For gasoline-fueled heavy-duty engines, the EPA presents a correction factor for NOx based on the humidity of the inlet air. This correction factor was established based on the work of Krause(3) in 1971. The vehicles were tested according to the Federal Heavy-Duty Test cycle and the resulting humidity correction can be calculated with the following equation:
Where:
G = Absolute humidity of the inlet air [grains H2O/pound dry air]
The promulgated correction is solely a function of the inlet air humidity. The original correction equation was a regression of the observed dependence of NO concentration in ppm on the inlet air humidity. The range of absolute humidities tested was from 20 to 110 grains/pound. Krause also established a correction equation for the mass emissions of NO2 g/bhp-hr as seen in the following equation:
Equations 2 and 3, in conjunction with NOx emissions modeled in Southwest Research Institute’s ALAMO_ENGINE cycle simulation computer model, have shown that correction factors established with concentration data and the corresponding mass-based emissions are similar. Therefore, a correction factor developed from concentration data imparts little error when applied to the mass-based emissions of an engine. Krause developed equations to correct carbon monoxide and unburned hydrocarbons emissions for ambient conditions, but the statistical correlation’s were not as strong as those from which equations 2 and 3 were derived.
The correction factors shown in equations 1-3 are used to adjust measurements of NOx to a reference humidity. The emission inventory models use correction factors in an inverse manner, to convert a standardized emissions rate to an actual rate. The equation and figures in the following sections are defined to adjust a standard emission rate to an actual rate based on ambient conditions.
3.1.2EPA Emission Inventory Models
The emissions inventory models used by EPA and CARB correct NOx emissions based on the average ambient conditions during engine operation. Emission models have been developed for both on-road and off-road mobile sources. The NOx correction factor used for a light-duty gasoline vehicle in EPA’s MOBILE6 is shown in the following equation(13) and Figure1 (equation is estimated from figure provided in EPA documents):
Where:
Ha = Absolute humidity of the inlet air [grains/lb]
Figure 1. Humidity Correction Factors as a function of Intake Air Absolute Humidity for Three Cases: MOBILE6 (equation 6), Reciprocal of Manos Light-duty Vehicle (Equation1), and Reciprocal of Krause Heavy-duty Vehicle (Equation 2).
Figure 1 shows three correction equations as a function of the inlet air humidity. While the empirical basis of the MOBILE6 function was not found, it appears to resemble the equation developed by Manos in the humidity ranges where the original regression analysis was performed. Continuing this speculation, it is probable that MOBILE6 model developers were not willing to extrapolate the Manos equation beyond the bounds of the regression analysis, and therefore, may not be accounting for the actual humidity effects on NOx formation at high humidity.
Temperature correction factors in MOBILE6 are determined separately for each of the three segments of the FTP for light-duty gasoline fueled vehicles. For ambient temperatures below 75-F the temperature correction factor (TCF) for NOx emissions is as follows(4):
Where:
TC(b) = Coefficient for the particular test segment
T = Ambient temperature [F]
TC(b) is dependent on the test segment, the ambient temperature, and the model year of the vehicle. At certain temperatures the TCF factor is combined with effects of fuel volatility as measured by the Reid vapor pressure. EPA’s NONROAD2002, the off-road emissions model, calculates the correction factors with the same algorithm as MOBILE6, but with a matrix of TC(b) specific to off-road, four-stroke engines. NONROAD2002 does not apply a correction factor to two-stroke engine emissions due to a lack of data for these engine types.
3.1.3CARB Emission Inventory Model
The CARB motor vehicle emission inventory model, EMFAC2002, corrects NOx emissions for the ambient conditions where the vehicle/engine operates. CARB based their humidity correction methodology on that published by Manos. CARB expands on the original methodology by adding a factor that is determined by the technology class for a given vehicle. The technology classes are differentiated by the method of fueling, such as multi-point fuel injection or carburetion, and the exhaust aftertreatment. The factors were created through a linear regression analysis of data recorded between 1989 to 1995 including 885 light-duty trucks, 116 medium-duty vehicles and 3447 passenger vehicles ranging in model 1962 to 1995(6). The equation developed to estimate NOx emissions for a vehicle operating at a humidity other than the standard 75 grains/pound takes the form:
Where:
Eamb = Corrected NOx mass emission
EStandard = NOx mass emissions at standard conditions
mmanos = -0.0047
mclass = ARB developed technology factors
HT = Absolute humidity during the vehicle/ test [grains/lb]
HS = Standard absolute humidity [grains/lb]
Hamb = Ambient absolute humidity during vehicle operation [grains/lb]
The mclass factors were obtained through linear regressions. None of the published statistical R2 values were larger than 0.033. Therefore, the adjusted correction has no statistical benefit over the original equation published by Manos.
EMFAC2002 corrects for the ambient temperature with technology specific correction factors for on-highway vehicles. The base equation takes the following form:
Where:
A, B and C = Technology specific coefficients
T = Ambient temperature [F]
Technology specific coefficients apply to the individual FTP segments and will adjust for engine differences such as fueling methods, exhaust aftertreatment, and air conditioning.
3.1.4Specialized Corrections
The effects of ambient conditions on the performance and emissions of two-stroke and four-stroke hand-held engines were explored by Brereton and Bertrand. They established a correction that takes the form:
Where:
AFR = Air-fuel ratio of the engine
= Absolute humidity of the inlet air [kg/kg]
This equation accommodates the range of AFR (A/F) that a hand-held carbureted engine may experience in actual use. Equation 8, with an operating A/F of 16, will reduce to the form defined by the EPA for use with small off-road engines.
3.1.5Comments on Current Correction Practices
Both the light-duty and heavy-duty correction equations, equations 1-3, were developed to correct for the minor variations in ambient humidity during a standardized test. The promulgated corrections are not suited for the wide range of humidity seen throughout the country. Use of any correction at conditions outside of the bound of the original regression should be considered an extrapolation. Equation 1, adopted by the EPA, did not include the effects of temperature because of the limited temperature range, and subsequent minimal effect on NOx concentration. If the temperature adjustment is not removed from Manos’ original equation to adjust an observed NOx emissions to a standard humidity, it takes the following form: